lMay5, 195s
CYCLICQUATERNARY SALTSAND
THEIR
2217
DECOMPOSITION
for 0.5 hour and the solution, after cooling and filtering, was made basic to phenolphthalein with 10 N sodium hydroxide. The crude product resulting was crystallized twice from isooctane (with charcoal decoloration); yield 20.0 g., m.p. 190.5191 .So, Xmnx 295 m p , El1 1184 in ethanol. ~ ~ Calcd. ~ for l c ~.H , ~ N ~ \c~,, o66.65; : H, 6.21; 3, 17.28. Found: C, 66.67; H , 6.29; N, 17.23, 17.32.
Anal. Calcd. for CloH19N2S: C, 63.12; H, 5.30; N, 14.73; S, 16.85. Found: C, 62.79, 62.74; H, 5.51, 5.54; N, 14.78; S , 16.62, 16.66. The infrared spectrum (KBr disk) exhibits a band a t 6.05 p characteristic of a carbon-carbon double bond and a bond a t 3.09 p for a N-H band. I n the ultraviolet the compound absorbs in three regions: Amax 279 m p , El1 1344; Xmax 240 3,4-Dihydro-3-methyl-4-methylene-2,1(H)quinazoline- m p y 663; Amax 338 in thione (Provisional Identification) .--4 solution of 10 g. of wish to thank hfr. o-aminoa~etophenone~~ (0.0i4 mole) and 5.4 g. of methyl isothiocyanate (0.074 mole) in 40 ml. of ethanol was reChilds and co-workers for the microfluxed 3 hours. The product, which began separating after analyses; hfr. Bruce Scott, Dr. J. hf.Vandenbelt approximately 1.5 hours, was collected and dried. A yield and co-workers for the ultraviolet and infrared deof 7.7 g. (50%) of crude product, m.P. 215-218', Jvas ob- terminations; and hlr. Richard Kavanaugh for tained. -4 sample for analysis, obtained by crystallization in the performance Of the in aivo successivelv from ethanol, 1-1 ethanol-Cellosolve and ethanol, melted a t 223-2260. tests. 2159
(38) N. J. Leonard a n d S. N. Boyd, Jr., J . Org. Chem., 11, 409 (1946)
[COSTRIBUTIOX FROM
Amine Oxides.
THE
DETROIT,MICHIGAN CHICAGO, ILLINOIS
DEPARTMENT O F CHEMISTRY, UNIVERSITY
ROCHESTER]
OF
Cyclic Quaternary Salts and their Decomposition' BY V. BOEKELHEIDE AND WAYNE FEELY~ RECEIVED DECEMBER 4, 1957
The synthesis of 2,3-dihydro-4H-oxazino[2,3-a] pyridinium bromide ( I V ) and its alkaline decomposition to 2-vinylpyridine bromide (VI) also is given but its and formaldehyde is described. The preparation of 2,3-dihydroisoxazolo[2,3-a]pyridinium alkaline decomposition yields products of unknown structure. The quaternization of 2-(P-bromoethyl)-pyridine results in dimerization and the product has structure X I rather than the simple azobicyclooctane structure previously assigned by Loffler.
Since the time of Meisenheimer's experiments on quaternary salts of amine oxidesI3i t has been known that the decomposition of such salts would yield aldehydes. -4lthough the yield of aldehyde obtained from such decompositions usually is poor, recent work has demonstrated t h a t under the proper circumstances the yield of aldehydes may be quite high and the reaction may have general usefulness in preparative work.4 A possible extension of this method would be the introduction of a carbonyl function into the side chain of simple pyridine derivatives. With this in mind we undertook to synthesize and study the properties of some simple cyclic quaternary salts of pyridine-N-oxide as shown by I. The synthesis of the simple salts corresponding to I, where n was 1 and 2, followed in a straightforward fashion. As shown below, the commercially available alcohols 2- (B-hydroxyethyl)pyridine (11) and 2- (7-hydroxypropyl)-pyridine (111) were oxidized to the corresponding N-oxides, these were converted to the bromides with hydrobromic acid, and the bromides readily cyclized to the desired quaternary salts. I n support of the assigned structures it was found that catalytic hydrogenation of 2,3-dihydro-4Hoxazino [2,3-a]pyridinium bromide (IV) over Adams catalyst gave the crystalline hydrobromide of 3-(2'-piperidyl)-propan-l-o1. This was identical in all respects with an authentic sample prepared by the hydrogenation of 111. Likewise, catalytic hydrogenation of 2,3-dihydroisoxazolo [2,3-aJpyri(1) This investigation was aided b y a grant from t h e National Science Foundation. (2) Union Carbide and Carbon Predoctoral Fellow, 1955-1956. (3) J. Meisenbeimer, A n n . , 397, 273 (1913). (4) W. Feely, W. L. Lehn a n d V. Boekelheide, J. O r g . Chem., 22, 1135 (1957).
dinium bromide (VI) gave the hydrobromide of 2-(2'-piperidy1)-ethanol, identical in all respects with a sample prepared by the hydrogenation of 11.
I
I
]$'hen these cyclic salts were treated with alkali, reaction occurred readily, but the products were not the expected pyridine aldehydes. Instead, the aqueous alkaline decomposition of IV gave 2vinylpyridine in 85% yie€d. If it is presumed that the first stage is the removal of a proton a t the 4-position to give V, it becomes understandable how the formation of 2-vinylpyridine could readily occur with accompanying elimination of formaldehyde. IVhen the aqueous solution from the reaction was steam distilled, formaldehyde was isolated from the distillate as its dimedon derivative.
IV
v
V.BOEKELHEIDE AND WAYNEFEELY
221s
I n the case of VI, treatment with aqueous alkali led to intractable tars. By analogy to the previous example, i t seemed likely that VI, after loss of a proton, would be converted to 2-vinylpyridineN-oxide (VII), a compound which would be expected to polymerize very readily. I n an attempt to establish the presence of 2-vinylpyridine-Noxide, the reaction of VI with aqueous piperidine was investigated. Under these conditions it was anticipated t h a t the addition of piperidine to 2vinylpyridine-N-oxide would successfully compete with the polymerization reaction. A crystalline product, m.p. 124-126”, separated directly from the reaction mixture and, as expected, its composition was in agreement with t h a t required for the adduct (VIII) of piperidine and 2-vinylpyridineN-oxide.
Vol. SO
methylpyridine, Sorm and Sedivy have shown that the quaternary salt is a dimer.’* Therefore, it seemed advisable to obtain definite evidence regarding the structures of Loffler’s compounds before undertaking the synthesis of I, where n = 0. NCHzCH2Br
I /b ‘ 1
+
,y3+) I/ I
v\/AY IX
XI
2Br-
4
H z Pt
XI1
i
I/
---+I
I1
When the “pyridoniuin” bromide, prepared as described by was subjected to hydrogenation over platinum, i t gave an oil, b.p. 100” a t 0.2 mm., having a composition corresponding to (C7H13N)x. The high boiling point indicated that the product was a dimer (x = 2) and a molecular weight determination confirmed this. Thus, the most reasonable structure for Loffler’s “pyridonium” bromide is X I and the structures assigned t o the various methyl analogs prepared in his studies7-I0 must be reconsidered. The behavior of XI toward base was of interest since Loffler had reported that such solutions turned red but did not yield 2-vinylpyridine, the expected Hofmann decomposition product. This observation, which we were able to confirm, indicated that the corresponding methine base possessed fair stability and suggested the possibility that X I might undergo base-catalyzed condensation with aldehydes to give X I I I . This would produce in a very simple way the carbon skeleton present in various lupin alkaloids. -4lthough this hope could not be realized, i t was found, during attempts to effect condensation using piperidine as catalyst, that XI underwent a smooth reaction with piperidine to give 6-(N-piperidino) -ethylpyridine (XIV). n‘hether 2-vinylpyridine is an intermediate or whether the reaction proceeds by a simple direct displacement was not determined.
However, a closer inspection of the physical properties and chemical behavior of the product melting a t 124-126’ makes the assignment of structure VI11 untenable. I t s ultraviolet absorption spectrum in alcohol shows a maximum a t 340 mp ( E 43,000) which is shifted irreversibly by acid to 248 mp ( E 15,000). Hydrogenation over platinum in ethyl acetate proceeded extremely rapidly with the uptake of four moles of hydrogen. Since the catalytic hydrogenation of VI11 would be expected to yield 2-( 1’-piperidyl)-ethylpiperidine, a sample of this was prepared by independent synthesis and by comparison was shown to be quite different. Bctually the oily hydrogenation product had a composition in accord with the enipirical formula ( C ~ H I ~ N Ocorresponding )~, to a loss of the piperidine moiety. With the evidence a t hand it is not yet possible to assign a satisfactory structure to this unusual product from the reaction of VI with piperidine.j The possibility of preparing salts such as I with smaller ring size, where n = 0, was considered as an alternate route for the preparation of 2-pyridine aldehyde. In the early work of Lijftler there is described the cyclization of 2-(@-haloethyl) XI11 pyridines (IX) to give pyridinum salts for which structure X was assigned.6 -1 number of such The ease with which X I underwent cleavage to compounds appear in the literature7-’0 and their structural assignments have been generally ac- give products related to 2-vinylpyridine suggested cepted even in recent publications.ll However, in that the reverse reaction might also occur. This the case of the next lower homolog, a-bromo- was found to be true. il’hen 2-vinylpyridine hydrobroinide was heated a t elevated temperatures, ( 5 ) Publication of these observations is included in t h e present i t gave XI in OS7, yield and this is probably the work becauze this research has been unavoidably interrupted. most convenient method for preparing X I . 1Vhether ( 6 ) K. L6FBer. Ber., 37, 161 (1904). 2-(r3-bromoethyl)-pyridineis a n actual intermedi(7) K. Loffler, ibid., 42, 948 (1909). (8) K. Loffler and A. Grosse, i b i d . , 40, 1325 (1907). ate in this case is uncertain. (9) K. Loffler and F. Thiel, ibid., 42, 132 (1909). (10) K . LoMer a n d P. Plocker, ibid., 40, 1310 (1907). ( 1 1 ) F. Bohlmann, N. Ottawa a n d R. Keller, Ann., 687, 102 (1954).
(12) F. sorni and L Sedivy, Coli. Czech. Cizetn. Cointn , 13, 289 (1948).
May 5, 1958
CYCLIC QUATERNARY SALTS AND
Experimental13 2,3-Dihydro-4H-oxazino [2,3-alpyridinium Bromide (IV). -A solution of 68.0 g. of 3-(2'-pyridy1)-propan-1-01-Noxide14in 1 1. of 48% hydrobromic acid was boiled under reflux for 24 hours. After concentration under reduced pressure, the residue was taken up in 1 1. of water and passed over an ion-exchange column (300 g. of Amberlite-IR-4BOH). Concentration of the eluate under reduced pressure gave a brown solid. This was crystallized from acetonitrile to give 55.0 g. (57%) of white crystals, m.p. 151-153'. Anal. Calcd. for C8HloNOBr: C, 44.46; H, 4.66. Found: C, 44.37; H, 4.86. 3-(2'-Piperidyl)-propan-l-o1 Hydrobromide. (A) By Hydrogenation of 1V.-A mixture containing 1.5 g. of 2,3dihydro-4H-oxazino [Z ,3-a]pyridinium bromide (IV) and 100 mg. of Adams catalyst in 75 ml. of water was subjected to hydrogenation a t room temperature and atmospheric pressure. Four moles of hydrogen was absorbed in 3 hours. After removal of the catalyst, the aqueous filtrate was concentrated under reduced pressure. The white solid residue, after recrystallization from acetone, gave in high yield white crystals, m.p. 126-128'. Anal. Calcd. for C8H18NOBr: C, 42.86; H, 8.04. Found: C, 43.16; H, 8.36. (B) By Hydrogenation of 3-( 2'-Pyridyl)-propan-l-ol Hydrobromide.-A solution of 1.0 g. of 3-(2'-pyridy1)-propan1-01 dissolved in 25 ml. of 5y0 hydrobromic acid was added to a suspension prepared by reducing 100 mg. of Adams catalyst in 20 ml. of water. The hydrogenation, conducted a t room temperature and atmospheric pressure, stopped when 3 moles of hydrogen had been absorbed. Evaporation of the solution under reduced pressure gave a white solid which, after recrystallization from acetone, gave white crystals, m.p. 125127'. A comparison of these crystals with a sample obtained as in (A), both by the method of mixed melting points and by use of their infrared spectra, confirmed their identity. Treatment of I V with Alkali.-To a solution of 10 g. of 2,3-dihydro-4H-oxazino 12,3-al~vridiniumbromide IIV) in 50 ml. 6f water was addgd 20 mi. of a 10% sodium hydioxide solution. The resulting turbid mixture was extracted with ether, the combined ether extracts were dried, and the ethereal solution was concentrated in a nitrogen atmosphere. Distillation of the residual oil gave 3.94 g. (85%) of a clear liquid, b.p. 68-70' a t 15 mm. The infrared and ultraviolet absorption spectra of this oil were identical with those of an authentic sample of 2-vinylpyridine. The picrate readily formed and, after recrystallization from ethanol, melted a t 157-159' (lit.6 gives 157-159'), undepressed by admixture of an authentic sample of the picrate of 2-vinylpyridine. The aqueous layer was steam distilled and the distillate was treated with 5,5-dimethyldihydroresorcinol (dimedon). The white crystals (12% yield) which separated melted a t 185-188O, undepressed by admixture of an authentic sample of the dimedon derivative of formaldehyde. 2-(2'-Pyridyl)-ethanol-N-oxidewas prepared in the same manner previously described for the preparation of 3-(2'pyridyl)-propan-l-ol-N-oxide.14 From 123.0 g. of 2-(2'pyridy1)-ethanol there was obtained, after recrystallization from dry ethyl acetate, 120.0 g. (86%) of white needles, m.p. 93-95', Anal. Calcd. for C7HpN02: C, 60.42; H , 6.52. Found: C, 60.51; H , 6.77. The picrate of 2-(2'-pyridy1)-ethanol-N-oxide formed readily and was obtained, after recrystallization from absolute ethanol, as yellow needles, m.p. 96-98'. Anal. Calcd. for C13H12N4O9: C, 42.40; H, 3.28. Found: C, 42.35; H, 3.55. 2,3-Dihydroisoxazolo [2,3-a]pyridinium Bromide (VI).A solution of 20.0 g. of 2-(2'-pyridyl)-ethanol-N-oxide in 500 ml. of 4870 hydrobromic acid was boiled under reflux for 10 hours. After the resulting solution had been concentrated under reduced pressure, the residual solid was taken up in 500 ml. of water and passed over an ion-exchange column (300 g. of Amberlite-IR-4B-OH). Concentration of the eluate gave a tan solid which, after recrystallization (13) All melting points are corrected. Analyses by Miss A. Smith and Micro-Tech Laboratories. (14) V. Boekelheide and W. Feely, J . OYE.Chem., '22, 589 (1957).
THEIR
DECOMPOSITION
2219
from acetonitrile, yielded 17.2 g. (60%) of white crystals, m.p. 152-155'. Anal. Calcd. for C,HsNOBr: C, 41.61; €I, 3.99. Found: C, 41.58; H , 3.92. The picrate of the 2,3-dihydroisoxazolo [2,3-a]pyridinium ion was prepared in ethanol and, after recrystallization from the same solvent, was isolated as yellow needles, m.p. 94-97", Anal. Calcd. for ClaHlzNdOs: C, 44.58; €1, 2.89. Found: C, 44.73; H , 3.04. 2-(2'-Piperidyl)-ethanol Hydrobromide. ( A ) By Hydrogenation of VI.--A solution of 1.2 g. of 2,3-dihyclroisoxazolo[2,3-a]pyridinium bromide in 30 ml. of water was added to a suspension of 100 mg. of prereduced Adams catalyst in 20 ml. of water. Hydrogenation was conducted a t room temperature and atmospheric pressure until 4 moles of hydrogen was absorbed (4.5 hours). Removal of the catalyst and evaporation of the filtrate gave a white solid which, after recrystallization from dry acetone, yieldeld white needles, m.p. 108-110". Anal. Calcd. for CTH16NOBr: C, 40.01; I-[, 7.68. Found: C, 40.13; H , 7.48. (B) By Hydrogenation of 2-(2'-Pyridy1)-ethanol Hydrobromide.-A solution of 1.0 g. of 2-(2'-pyridy1)-ethanol in 25 ml. of 5y0 hydrobromic acid was added to a suspension of 100 mg. of prereduced Adams catalyst in 20 ml. of water. After hydrogenation a t room temperature and atmospheric pressure, the catalyst and solvent were removed. The residual solid, when recrystallized from acetone, gave 1.0 g. of white needles, m.p. 108-110'. A comparison of the samples from (A) and (B) by infrared spectra and the method of mixed melting points showed the two to be identical. Reaction of VI with Piperidine.-A solution of 10.0 g. of 2,3-dihydroisoxazolo [2,3-a]pyridinium bromide (VI) in 100 ml. of water was added to 100 ml. of piperidine. The resulting homogeneous solution was warmed on a ste:im-bath for 15 minutes and then allowed to cool slowly. The yellow crystals, which separated from the solution, were collected and recrystallized from ethaaol to give 4.45 g. (42%) of The ultraviolet absorption yellow plates, m.p. 124-126 spectrum of the crystals in ethanol showed maxima a t 238 mp (e 2,900) and 340 m p (e 43,000). When aqueous acid was added to the solution, a single maximum was observed a t 248 mG (e 15,000). Addition of base left the spectrum unaffected and did not regenerate the original spectrum observed in neutral solution. While the crystals could be dissolved in alcohol without change, their benzene and chloroform solutions showed rapid decomposition on warming. Anal. Calcd. for C12HlaN20: C, 69.89; H , 8.30; N, 13.58. Found: C, 69.93; H , 8.89; N, 13.57. When a solution of 2.0 g. of these crystals in 20 ml. of ethyl acetate was subjected to hydrogenation over 100 mg. of Adams catalyst a t room temperature and atmospheric pressure, 4 moles of hydrogen was absorbed very rapidly. Removal of the catalyst and solvent and then distillation of the residual gum gave a colorless, viscous oil, b.p. 130" a t 1 mm. The high boiling point of this oil would suggest that it is a dimer (Le., x = 2). Anal. Calcd. for (C7H16NO),: C, 65.07; H , 11.70; N, 10.84. Found: C, 65.17; H, 11.88; N, 10.52. 3,4,7,8-Tetrahydrodibenzo[a,e]-1 ,S-diazocinium Dibrosolution of 12.3 g. of 2-(2'-pyridy1)-ethanol in mide (=).-A 150 ml. of 48y0 hydrobromic acid was boiled under reflux for 12 hours. After concentration, the white crystalline solid, 2-(0-bromoethy1)-pyridine hydrobromide, was taken up in 100 ml. of water, neutralized with sodium carbonate, and extracted with ether. The combined ether extracts were dried and concentrated to give a white solid which, .ifter recrystallization from ethanol-water, yielded 13.5 g. (73%) of white needles, m.p. 238-240°.16 Anal. Calcd. for CIIHIBNZBrZ: C, 45.18; H , 4.54. Found: C, 44.98; H , 4.41. The dipicrate of X I was prepared in aqueous solution and was obtained, after recrystallization from water, as yellow needles, m.p. 179-181'. Anal. Calcd. for CzsHzoN801,: C, 46.71; H , 3.02. Found: C, 46.i6; H , 3.35.
.
(15) L(iffler8 gives 226-227'.
2220
BARNEYJ. MAGERLEIN AND J O H N A,HOGG
The diiodide of X I was prepared by adding potassium iodide t o an aqueous solution of X I . After recrystallization from water it gave pale yellow crystals, m.p. 230-231°.10 Anal. Calcd. for C19H16hT212: C, 36.07; H, 3.46. Found: C, 36.09; H , 3.76. When an attempt was made to oxidize X I with dichromate in the hope of preparing the parent substance XV, the only reaction was the precipitation of the dichromate salt of XI. This, after recrystallization from water, yielded orange needles, m.p. 170-175" dec. Anal. Calcd. for Cl4HiBN20&r2: C, 39.26; H , 3 . 7 . Found: C, 39.05; H , 3.89. Similarly, attempts t o oxidize XI using bromine water simply precipitated an unstable, yellow di-perbromide.
1,Z,5,6-Bis-( tetramethy1ene)-octahydro-1,S-diazocine (XII).-A solution of 1.60 g. of 3,4,7&tetrahydrodibenzo[a,e]-1,5-diazocinium bromide (XI) in 25 ml. of water was subjected to hydrogenation a t room temperature and atmospheric pressure using 100 mg. of Adams catalyst. When 6 moles of hydrogen had been absorbed, the hydrogenation was stopped and the catalyst removed. Neutralization of the filtrate followed by extraction with chloroform and concentration gave 600 mg. of a colorless oil; b.p. 105' a t 0.2 mm., nZ4D 1.5800. The infrared spectrum of the oil showed a strong band a t 3.63 p in addition to the usual absorption a t 3.45 p . As has been pointed out by Wenkert and Roychaudhuri," strong absorption in this region probably is due to an axial hydrogen a t the bridgehead carbon. Anal. Calcd. for ClrH2~Nn:C, 75.61; H , 11.79; mol. wt., 220. Found: C, 75.25; H, 11.58; mol. wt. (Rast), 208. The dimethiodide of XI1 formed readily and, after recrystallization from a water-ethanol mixture, was obtained a s white needles, m.p. 282-283' dec. Anal. Calcd. for C I B H ~ J V ~C, I Z38.26; : H, 5.62. Found: C, 37.87; H , 5.61. Reaction of XI with Piperidine.-To a solution of 2.0 g. of X I in 4 ml. of water was added 1 ml. of piperidine. After the mixture had been boiled under reflux for 3 hours, it was concentrated under reduced pressure t o give a tan solid. This, after recrystallization from a benzene-ethanol mixture, yielded 2.9 g. (89%) of N-(P-(2-pyridyl)-ethyl)-piperidine hydrobromide a s white crystals, m.p. 173-175'. Anal. Calcd. for CI2Hl9N2Br: C, 53.14; H , 7.06. Found: C, 53.99; H, 7.45. Conversion of the above crystals to the corresponding di(16) Lofflers gives 211-213'. (17) E. Wenkert and D . IC. Roychaudhuri, THISJ O U R N A L , 7 8 , 6417 (1956).
Vol.
so
picrate gave yellow crystals, m.p. 160-161.5°.18 A mixture melting point determination with an authentic sample of the dipicrate of X-(6-(2-pyridyl)-ethq-l)-piperidineshowed no depression of melting point. N-(2-(2'-Piperidyl)-ethyl)-piperidine.--A solution of 19.0 g. of N-(P-(2-pyridyl)-ethyl)-piperidine18 in 200 ml. of a 10% hydrobromic acid solution containing 500 mg. of Adams catalyst was subjected t o hydrogenation a t room temperature and atmospheric pressure. After 3 moles of hydrogen had been absorbed, the catalyst was removed and the solution was made basic. Extraction with ether followed by concentration and distillation of the residue gave 18.0 g. (93%) of a colorless oil, b.p. 107' a t 1 mm., ? t Z 9 ~ 1.4886. This was prepared for comparison with the reduction product from VI and was shown by comparison of spectra and other physical properties to be different. Anal. Calcd. for C12H21N2:C, 73.41; H, 12.33. Fouiiti: C, 73 51; H , 11.90. Pre ration of XI from 2-Vinylpyridine Hydrobromide.-1 sample of 4.0 g. of 2-vinylpyridine hydrobromide was heated in a sealed tube a t 160" for 3 hours. The resulting solid was washed from the tube with absolute ethanol and collected by filtration. After recrystallization from an ethanol-water mixture, it yielded 2.7 g. (68%) of crystals, 9 comparison of these crystals with the m.p. 238-240'. sample of X I previously prepared showed them t o have identical infrared spectra and a mixture of the two showed no depression of melting point. 3 ,&Dihydrodibenzo[a,d]-1,4-diazinium Dichloride (XV). -In studying the behavior of X I toward_alkali we were interested in repeating the observations of Sorm and Sedivy12 on the analogous system XVI for comparison. When a solution of 2-chloromethylpyridine in dry acetonitrile was boiled under reflux for 5 days. a white solid separated. This, after recrystallization from an ethanol-water mixture, gave white needles, m.p. 253-255'. Anal. Calcd. for Cl2H12N2C12: C, 56.48; H, 4.74. Found: C, 56.81; H, 5.02. When an aqueous solution of these crystals was treated with dilute alkali, an intense red color developed and the colored substance could not be extracted from the aqueous solution with organic solvents. However, with stronger alkali, a highly unstable, dark red compound, soluble in organic solvents, did separate. I t would appear that the intense red color formed in dilute alkali is due to XVII and only in strong alkali is the non-ionic, free base produced.
XVI
2 c1-
XVII
c1-
(18) W. E. Doering and R. A. N. Weil, h i d . , 69, 2461 (1947), give 159-160° as t h e melting point of the dipicrate of N-(8-(2-pyridyl)ethy1)piperidine.
ROCHESTER, NEWYORK
[CONTRIBUTIOS FROM THE RESEARCH LADORATORIES, THE U P J O H N C O . ]
Preparation and Reactions of 11-Substituted 1,3,5(lO)-Estratrienes,l genated Estrones and Estradiols U Y UAKNEYJ. MAGERLEIN AND
JOHN
I.
11-Oxy-
A . Ilotic;
RECEIVED DECEMBER 26, 1957 'l'lie arotriatizatioii of steroidal 3-keto-l,4dieiies containing lla-hydroxy, 1la-acetoxy ;iild 11,l-hytiroxy stilxtituciith 01' !l,ll-utisaturation is reported. The conversion of these products to the 11-oxygenated 19-nortestosteroncs is the first reported synthesis of this type of compound wherein the 11-hydroxyl group is present during the chemical modification of ring A . A novel compound obtained by a Birch-type reduction is described.
The pyrolytic method for the aromatization of has been used for the preparation of a variety of 3-keto-] ,.l-dienes announced by Inhoffen in 1940,2 aromatic A ring steroids unsubstituted at C-I The pyrolysis Of ll~B-hydrox>',lla-hydroxy, ' l o (1) A preliminary announcement of this work was made i n a Communicntion t o t h e Editor, T H I S J o U R N A r . , 1 9 , 1508 (19j7). ( 2 ) €1. 15. InhoBen. . 4 n r ~ l u .C h m . , SS, 471 (1940).
r.
( 3 ) See U . J. lI;1gerleiri : ~ n d A . I I o K ~ ?'rlv(ihcdroii , ( i n p r c \ < ) , for a I,iicf irvicw of (hi:: rcxrtion.
